1994.08.09 11:40 "Is TIFF suitable for 3D surface data ?", by Wolfram Gloger

Hello,

A while ago I posted the following to comp.graphics, but got no response. So I assume that all experts are gathered on this list. Note: I am _not_ subscribed.

We are currently trying to decide on a suitable (but as standardized as possible) format for surface data in 3D space. The data is given by an MxN array of pixels, with each pixel consisting of x, y and z values (in floating point). The 'grid' defined by the x and y values is 'almost' (but not exactly) regular, i.e. differences between x/y values with adjacent indices are 'almost' constant.

This gives some justification for a quick viewing program of these files to forget about the x and y and only display the z values (mapped to gray-scale, for example).

I am thinking about using TIFF, setting tags as follows:

ImageWidth N
ImageLength M
SamplesPerPixel 3
BitsPerSample 32,32,32
SampleFormat 3,3,3 (SampleFormat_IEEEFP)

Is there something wrong with this---maybe there's a better way with TIFF? But suggestions for another suitable raster format would also be very welcome.

Thanks,

Wolfram Gloger
(u7y22ab@sunmail.lrz-muenchen.de, wmglo@dent.med.uni-muenchen.de)